TL;DR
It is concluded that only by grounding oversight in realistic assumptions about human cognition and institutional incentives can it fulfill its intended role in ensuring trustworthy AI.
Human oversight is widely held to mitigate the risks of AI systems. Even for systems that produce discrete outputs at identifiable decision points, the realisation of human oversight as a reactive measure is empirically fragile, yet increasingly well understood. However, for agentic AI -- systems that plan, decompose g...
The Stoic concept of the hegemonikon—the capacity for assent that directs the mind—is used as a framework for understanding intellectual creation and decision-making, clarifying when AI supports human judgment and when it replaces decision-making authority.
Human-in-the-loop (HITL) approaches are commonly proposed to address alignment challenges arising from the use of large language models in research ethics oversight. This paper argues that, paradoxically, a HITL can amplify the risk of misalignment. Using the example of protocol triage in research ethics oversight, I d...
It is found that at least for the time being, explicit normative instructions are not fully able to realign AI advice with the normative convictions of the population, or the legislator deciding on its behalf.
The construct of calculated inefficiency is developed, defined as the structural property of a hybrid information system in which human interventions in automatable processes are maintained at levels above what pure efficiency optimisation would select, regardless of whether that maintenance was deliberately architecte...
It is argued that demonstrating internal incoherence is a necessary precursor to AI alignment as well as a broader phenomenon of epistemic instability in generative AI wherein models fail to reliably maintain coherence with respect to their own prior outputs.
We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.
